Basics of Solar Hydrogen

thoughtgreenpepperΜηχανική

27 Οκτ 2013 (πριν από 3 χρόνια και 9 μήνες)

88 εμφανίσεις

Tfy
-
56.5141 Special Course in Advanced Energy Technologies


Erno Kemppainen

1


Basics of Solar Hydrogen

Photoelectrochemical water splitting

(PWS)

is a process where a semiconductor material absorbs
photons and channels their

energy to decomposing water molecules into hydrogen and oxygen

[1]
.
Hydrogen can be used as such in fuel cells and its energy density is very high, so it is
a
potentially
excellent energy storage even if there are problems with storing it.

Therefore PWS can potentially
increase the amount of solar energy production witho
ut the problems with the electric grid control
that are common with many renewable energy sources.

In order to produce hydrogen from water with sunlight as the only energy source, the
conduction
band level

of

the semiconductor must
be more negative than th
e thermodynamic potential of the
hydrogen evolution reaction (HER) and the valence band must be more positive than the potential
of oxygen evolution reaction (OER)

[1, 2]
. (See figure 1.) Therefore the minimum bandgap of the
semiconductor is 1.23 eV, if on
ly one material is used to split the water molecules

[2]
. It is also
possible to use a device configuration with several different materials, in which case smaller
individual bandgaps can be used

[1]
. Additionally, the material(s) must be stable in the ope
rating
conditions and absorb light efficiently. Good reaction kinetics on the material surfaces (low
overpotentials) are almost a requirement for efficient light absorption, because high overpotentials
would require large bandgap for reasonable hydrogen pr
oduction rate. Consequently, large bandgap
would reduce the amount of photons that could be absorbed and therefore limit the hydrogen
production rate (Figure 1.). Typically a bandgap of about 2.0 eV is considered as an optimal case,
when considering the co
mbined effects of light absorption and overpotentials

[1]
. Low price and
abundant supply would also be advantageous for mass production of PWS equipment.


Figure 1.

AM1.5G spectrum and the bandgap positions of var
ious metal oxide semiconductors
(left) and the
e
ner
gy band positions for various semiconductors at pH14

(right).

[1, 3]


Overpotentials are voltage losses caused by reaction kinetics that require some finite potential
difference to maintain a nonzero net reaction rate (electric current). The higher the
rate, the higher
the overpotential. The overpotential related to HER (H
+
/H
2

in figure 1.) is typically small compared
to the overpotential of OER (OH
-
/O
2
), at least in part because more electrons are transferred in the
OER than in the HER. The problem with

OER overpotential is that it might be impossible to
develop a catalyst that would remove this overpotential entirely and there may always be significant
energy losses related to this half reaction.

[2]

The main issue with the PWS is the lack of materials
that would both be stable and produce
hydrogen efficiently
. Typically only metal oxide semiconductors are stable in the operating
conditions; most other semiconductors decompose or form a thin oxide layer that prevents electron
Tfy
-
56.5141 Special Course in Advanced Energy Technologies


Erno Kemppainen

2


transfer across the semicond
uctor/liquid
-
interface

[1]
. However, metal oxides are not
thermodynamically stable and survive only, because their charge transfer kinetics across the
interface are faster than the photocorrosion reaction

[1]
. Additionally, large bandgap does not
directly
indicate stability: TiO
2

and hematite (
α
-
Fe
2
O
3
) are stable in typical operating conditions,
but ZnO with larger bandgap decomposes, because its charge transfer is not rapid enough

[1]
. (See
figure 1.)

Hematite and TiO
2

are very common materials in water sp
litting research. They are both cheap
stable and abundant, but also have their own issues with the energy levels: The conduction band of
hematite is too positive for HER and the bandgap of TiO
2

is too large for visible light absorption

[1,
4, 5]
. One chara
cteristic common for both materials is the short valence band hole diffusion length
compared to the material thickness required for efficient light absorption

[1, 4, 6]
. This necessitates
a porous nanostructure to enable efficient charge separation, while
maintaining the film thickness
required for efficient light absorption. Additional benefits of such structures include the fact that the
bandgap of nanoscale hematite is slightly larger than that of bulk hematite, which indicates that the
conduction band o
f hematite in these materials may be negative enough to enable HER

[6]
. Even if
the charge separation would not require nanostructured films, the increased surface area would
enhance the hydrogen production by providing the surface reactions with more adso
rption and
reaction sites than a smooth crystal surface.

Another development approach typical for all materials is the control of the surface of the
semiconductor to either introduce catalyst particles to reduce the overpotentials or for example
plasmonic
nanoparticles to enhance light absorption. It has been shown that plasmonic metal
nanoparticles typically enhance light absorption so that light can be absorbed efficiently in thinner
films

[7]
. Also the absorption of longer wavelengths is often improved

[
7]
. It is also possible that
the improved light absorption does not transfer into increased hydrogen production, because of the
reduced surface area, recombination and other detrimental effects

[8]
. Therefore careful design and
optimization is required for

fully utilizing the absorption enhancement of plasmonic nanoparticles.


Figure 2.

Absorption

spectra

of TiO
2

(a) with

different concentrations of

(A) implanted Cr ions
and (B) chemically doped with Cr ions
[5]

As TiO
2

is a cheap, abundant and stable material and i
t is possible to manipulate its energy levels
with doping,
much research is centred on modifying
its
electronic structure.
The doping can be done
in several ways, but chemical solution methods and other surface doping methods typically yield
only modest im
provements in light absorption with insignificant increases in photocurrent

[5]
. On
the other hand, methods that can introduce dopants into the bulk of the nanoparticles, such as ion
implantation, have been shown to create band
-
like states that enable visi
ble light absorption and
increase the total catalytic activity of the doped materials

[5]
. (See figure 2.) Common dopant
Tfy
-
56.5141 Special Course in Advanced Energy Technologies


Erno Kemppainen

3


materials include for instance carbon (C), chromium (Cr), platinum (Pt) and nitrogen (N)

[5, 9, 10]
.
So far, these doped TiO
2

films hav
e required additional voltage or pH bias to successfully
decompose water molecules into hydrogen and oxygen, but they have accomplished that under
visible light illumination, unlike undoped TiO
2

[5]
.

References


[1]

R. van de Krol, Y. Liang and J. Schoonman, "Solar hydrogen production with nanostructured metal
oxides,"
Journal of Materials Chemistry,
vol. 18, pp. 2311
-
2320, 2008.

[2]

M. T. Koper, "Thermodynamic theory of multi
-
electron transfer reactions: Implicati
ons,"
Journal of
Electroanalytical Chemistry,
vol. 660, pp. 254
-
260, 2011.

[3]

J. Sun, D. K. Zhong and D. R. Gamelin, "Composite photoanodes for photoelectrochemical solar water
splitting,"
Energy & Environmental Science,
vol. 3, p. 1252

12S1, 2M1MK

[
4崠

K. Sivula, F. Le Formal and M. Grätzel, "Solar Water Splitting: Progress Using Hematite (α
-
Fe2lP⤬"
ChemSusChem,
vol. 4, pp. 432
-
449, 2011.

[5]

M. Takeuchi, M. Matsuoka and M. Anpo, "Ion engineering techniques for the preparation of the highly
effec
tive TiO2 photocatalysts operating under visible light irradiation,"
Research on Chemical
Intermediates,
vol. 38, pp. 1261
-
1277, 2012.

[6]

V. A. de Carvalho, R. A. d. S. Luz, B. H. Lima, F. N. Crespilho, E. R. Leite and F. L. Souza, "Highly
oriented hem
atite nanorods arrays for photoelectrochemical water splitting,"
Jounal of Power Sources,
vol. 205, pp. 525
-
529, 2012.

[7]

S. C. Warren and E. Thimsen, "Plasmonic Solar Water Splitting,"
Energy & Environmental Science,
vol. 5, p. 5133

514S, 2M12K

嬸崠

AK saädes, gK 䉲Bääet, MK d狤tzeä, eK dudmundsdotti爬 eK eansen, eK gonsson, mK Käup晥ä, dK K牯es, FK
ie Fo牭aä, 䤮 Man, 刮 Ma牴ins, gK No牳kov, gK 副ssmeisä, KK pivuäa, AK sojvodic and MK Zach, "poäa爠
hyd牯gen p牯duction with semiconducto爠metaä oxidesW n
ew di牥ctions in expe物ment and theo特,"
Physical chemistry chemical physics,
vol. 14, pp. 49
-
70, 2011.

[9]

Y. A. Shaban and S. U. Khan, "Visible light active carbon modified n
-
TiO2 for efficient hydrogen
production by photoelectrochemical splitting of
water,"
Hydrogen Energy,
vol. 33, pp. 1118
-
1126,
2008.

[10]

M. Kitano, M. Takeuchi, M. Matsuoka, J. M. Thomas and M. Anpo, "Photocatalytic water splitting
using Pt
-
loaded visible light responsive TiO2 thin flim photocatalysts,"
Catalysis Today,
vol. 120
, pp.
133
-
138, 2007.